An image sensor typically includes an array of pixels. When an image of a scene is to be captured by the image sensor, each pixel accumulates photo-generated charge based on the amount of light striking the pixel. Column circuits connected to the pixels receive voltage signals from the pixels and convert the voltage signals into digital signals using analog-to-digital converters (ADCs) included in each column circuit. The digital signals are then combined to produce the captured image.
Single slope ADCs are used in many image sensors due at least in part to their linearity and to their limited power consumption. However, for some image sensors, such as large pixel array image sensors, the single slope ADCs can take a relatively long time to convert all of the voltage signals into digital signals. Several readout techniques have been used to reduce the conversion time of the single slop ADCs. One technique places column circuits on two sides of the pixel array (e.g., top and bottom) to improve the readout throughput. Some of the pixels are readout by the column circuits on one side of the pixel array while the remaining pixels are readout by the column circuits on the other side of the pixel array.
Another technique uses multiple banks or groups of column circuits and staggers the use of the column circuits with respect to time. One group of column circuits reads the voltage signals from the pixels while another group of column circuits converts the voltage signals into digital signals. However, both of these techniques can increase the column fixed pattern noise in the digital signals, consume a greater amount of die area, and increase the amount of power consumed by the column circuits.